Statistical saturation attack
An attack that pins part of the plaintext, lets the rest vary, and measures how lopsided a few output bits become.
Baudoin Collard and François-Xavier Standaert introduced it in 2009 and aimed it at reduced-round versions of the lightweight cipher PRESENT. The idea borrows from integral attacks: hold some plaintext bits fixed and run the others through many values. Rather than looking for exact balance, the attacker measures statistically how evenly a handful of ciphertext bits are distributed. If the permutation layer diffuses poorly, the fixed bits leak through more rounds than the designers budgeted for.
Gregor Leander showed in 2011 that the technique is fundamentally multidimensional linear cryptanalysis viewed from the other end. That didn't make it less useful, but it did move it from its own shelf to a branch of an older tree.